Every few weeks someone asks why Farna Control is not built on Temporal, n8n, or LangGraph. All three orchestrate things, but workflow automation and agent orchestration serve different purposes.

What workflow tools do

Workflow tools (Temporal, n8n, Airflow, Prefect, LangGraph) handle task sequencing and state management. They answer:

Did step A complete before step B started?
If step C fails, should we retry or abort?
What is the current state of this long-running process?
How do we distribute work across workers?

A workflow tool does not know whether step A is "run a linter" or "charge a credit card" - only that A must finish before B starts.

What agent orchestration does

Agent orchestration answers a different set of questions:

Did the agent produce output that matches the expected schema?
Should this output be reviewed by a human before it reaches production?
What evidence supports the agent's decisions?
How do we know this agent is ready for more autonomy?
What happened in every run, and can we reproduce it?

These are governance questions about AI agents producing production work.

The practical difference

An agent writes a pull request. A workflow tool can trigger when the PR is opened, run the agent, wait for it to finish, and post the result back. An agent orchestrator can:

Trigger when the PR is opened
Run the agent in an isolated sandbox
Validate the agent's output against a schema
Run deterministic gates (regex, schema check, scored eval)
If a gate fails, spawn a correction run with the gate's evidence attached
If all gates pass, present the evidence to a human reviewer
Record every decision in an append-only audit trail
Record how often each gate fires, so you have evidence when you decide to raise a pipeline's autonomy level

Why Farna Control is not built on a workflow tool

Farna Control's pipeline model looks like a workflow on the surface - nodes connected by edges, triggers, conditions - but each node is a governance primitive: sandbox agent execution, schema validation, deterministic gates, HITL review, audit recording, bounded autonomy.

Building Farna Control on Temporal or n8n would mean implementing every governance primitive as a custom extension on a generic engine: a less coherent system with more surface area for bugs and a harder time providing what governance requires: append-only records, reproducible snapshots, evidence attachment.

For a detailed comparison of Farna Control against a specific general automation tool, see Farna Control vs n8n. For the broader self-hosted SDLC picture, see the self-hosted agent governance guide.

The principle behind this: deterministic gates catch failures before they reach a human reviewer, and audit as a first-class output means the history you inspect is exactly what happened.